Thrombin and Fibrinolysis*
Michael Nesheim, PhD
When the activities of the coagulation and fibrino-
lytic cascades are properly regulated, so that fibrin
(FN) deposition and removal are properly balanced,
the vascular system is protected from catastrophic
blood loss at the site of an injury, while its fluidity is
ensured elsewhere. When these activities are not
properly regulated, however, the organism is sub-
jected to either excessive bleeding or thrombosis.
Thrombomodulin on the endothelial cell is very
important in this regulation because it converts
thrombin to an anticoagulant enzyme by directing it
toward the activation of protein C. It also converts
thrombin to an antifibrinolytic enzyme by directing
it toward the activation of thrombin-activatable fibri-
nolysis inhibitor (TAFI). By doing so, it creates a
direct molecular connection between the coagula-
tion and fibrinolytic cascades, such that activation of
the former suppresses the activity of the latter.
Recent studies indicate that the TAFI pathway func-
tions in vivo and is likely relevant in maintaining the
proper balance between FN deposition and removal.
Whether it will be a target for pharmaceutical ma-
nipulation of this balance remains to be determined.
(CHEST 2003; 124:33S–39S)
Key words: bleeding; coagulation; fibrinolysis; plasmin; throm-
bin; thrombin-activatable fibrinolysis inhibitor; thrombomodulin;
thrombosis
Abbreviations: FDP ⫽ fibrin degradation product; FN ⫽ fibrin;
FPA ⫽ fibrinopeptide A; FPB ⫽ fibrinopeptide B; GPg ⫽ glu-
plasminogen; GPn ⫽ glu-plasmin; LPg ⫽ lys-plasminogen; LPn ⫽
lys-plasmin; PAI-1 ⫽ plasminogen activator inhibitor type 1;
PTCI ⫽ potato tuber carboxypeptidase inhibitor; TAFI ⫽ thrombin-
activatable fibrinolysis inhibitor; TAFIa ⫽ carboxypeptidase B-like en-
zyme; tPA ⫽ tissue-type plasminogen activator
T protects
he ongoing deposition and removal of fibrin (FN) both
the organism from untoward blood loss at the
site of injury and maintains the fluidity of the blood
elsewhere.1– 4 FN deposition occurs on activation of the
coagulation cascade. This results in the ultimate conver-
sion of prothrombin to thrombin, which then catalyzes
conversion of soluble fibrinogen to the gel-like substance,
FN, the major proteinaceous component of the clot
(Fig 1). FN removal occurs on activation of the fibrinolytic
cascade. Plasminogen then converts to the enzyme plas-
min, which catalyzes digestion of the FN clot to soluble
FN degradation products.
The activities of the cascades are normally latent. Their
*From the Departments of Biochemistry and Medicine, Queen’s
University, Kingston, ON, Canada.
Reproduction of this article is prohibited without written permis-
sion from the American College of Chest Physicians (e-mail:
permissions@chestnet.org).
Correspondence to: Michael Nesheim, PhD, Departments of
Biochemistry and Medicine, Botterell Hall, Room A210, Queen’s
University, Kingston, ON, Canada, K7L 3N6; e-mail: nesheimm@
post.queensu.ca
www.chestjournal.org
Figure 1. The balance between the formation and degradation
of FN. The coagulation cascade ultimately generates thrombin,
which catalyzes the conversion of fibrinogen to the fibrin clot.
The fibrinolytic cascade generates plasmin, which catalyzes sol-
ubilization of the FN. The thrombin-thrombomodulin complex
promotes down-regulation of thrombin formation by generating
activated protein C (APC). It also suppresses fibrinolysis by
forming TAFIa. The two cascades are thereby linked through the
thrombin, thrombomodulin, and TAFI pathway. PC ⫽ protein C.
potentials, however, have been shown in an animal model
(chimpanzee) to be remarkable.5 The coagulation cascade,
for example, when acutely stimulated, can convert the
entire plasma content of fibrinogen to FN and force the
level of circulating platelets to zero in less than a minute.
This potentially lethal phenomenon, however, is matched
by an equally potent fibrinolytic response, such that within
a minute or two, the plasma level of endogenous tissue
plasminogen activator increases several hundred fold, the
deposited FN is completely solubilized, and the circulat-
ing platelet level returns to normal. All of this occurs
without long-term negative effects on the animal. Thus,
both the coagulation and fibrinolytic cascades have the
potential to generate very profound activities in a very
acute manner. When the processes of FN deposition and
removal are properly regulated, they perform their phys-
iologic functions remarkably well. When they are not
properly balanced, however, the pathophysiologic conse-
quences are reflected in tendencies to bleed or throm-
bose. Prevalent thrombotic events include heart attacks
and strokes.
The regulation of the function of the cascades is
expressed through the endothelium, the blood platelets,
and the coagulation and fibrinolytic plasma proteins. One
key player, indicated in Figure 1, is thrombomodulin,6 an
integral membrane protein found on the endothelial cell.
It binds to thrombin and changes its substrate specificity,
such that it no longer recognizes fibrinogen as a substrate.
Instead, it catalyzes conversion of the zymogen, protein C,
to the anticoagulant enzyme, activated-protein C. This
enzyme inactivates factors Va and VIIIa of the coagulation
CHEST / 124 / 3 / SEPTEMBER, 2003 SUPPLEMENT 33S
cascade and thereby down-regulates thrombin formation.
This subject is discussed in detail in the article by Dr.
Esmon in this supplement. In addition, thrombin, partic-
ularly when it is bound to thrombomodulin, catalyzes the
activation of the zymogen thrombin-activatable fibrinolysis
inhibitor (TAFI) to the antifibrinolytic, carboxypeptidase
B-like enzyme (TAFIa),7,8 which down-regulates the co-
factor activity of FN in plasminogen activation and thereby
suppresses fibrinolysis.9 The pathway defined by throm-
bin, thrombomodulin, and TAFI creates a direct molecu-
lar connection between the coagulation and fibrinolytic
cascades, such that activation of the former suppresses the
activity of the latter.10 Thus, this pathway may play a key
role in the regulation of the balance between FN deposi-
tion and removal.
The Fibrinolytic Cascade
The formation of FN from fibrinogen is depicted in
Figure 2. The fibrinogen molecule comprises two globular
D domains and a central, smaller, globular E domain.
Thrombin catalyzes removal of fibrinopeptide A (FPA)
and fibrinopeptide B (FPB) from the molecule. These
events expose polymerization sites within the E domain
such that E domains associate specifically, tightly, and
noncovalently with sites on the D domains of neighboring
molecules to initiate the polymerization. Protofibrils con-
Figure 2. The conversion of fibrinogen to FN. When thrombin
acts on the soluble fibrinogen monomer, it catalyzes release of
FPA and FPB from the amino termini of the ␣ and ␤ chains of
fibrinogen. This exposes polymerization sites in the E domains of
fibrinogen such that noncovalent associations occur between the
E and D domains of neighboring molecules, thereby producing
double-stranded protofibrils. These laterally associate to form
bundles, which continue to grow and branch, ultimately forming
a three-dimensional web, which causes the solution to gel,
thereby forming the familiar blood clot. The web is stabilized by
covalent crosslinkages between adjacent D domains, which are
formed by factor XIIIa. A carboxypeptidase B (CPB)-like enzyme
also removes the COOH terminal arginine from FPB.
34S
sisting of two linear polymers form in a half-staggered
arrangement. Simultaneously, factor XIII is activated to
factor XIIIa, and it catalyzes covalent isopeptide bonds
between ␥ chains in the D domains of adjacent FN
monomers within the growing polymer. With further
liberation of FPA and FPB, double-stranded protomers
laterally associate to make FN bundles, which extend in
length and occasionally form branches. As a result, a FN
mesh develops that resembles a “three-dimensional spider
web.” The solution in which this occurs becomes gel-like
and forms, along with entrained cells, the familiar blood
clot. In addition, a carboxyl terminal arginine is removed
from FPB by a carboxypeptidase B-like enzyme. The FN
clot is subsequently digested when the fibrinolytic cascade
is triggered. These events are depicted in Figure 3.11
When FN forms, it serves as a cofactor for the conversion
of glu-plasminogen (GPg) to glu-plasmin (GPn). This
process begins digestion of FN by catalyzing cleavage after
specific arginine and lysine residues in the ␣, ␤, and ␥
chains of FN. The FN, while still clotted, is modified to a
form designated FN⬘ in Figure 3. This modified form has
in its structure carboxyl terminal lysine residues, which
promote the binding of tissue-type plasminogen activator
(tPA) and GPg to FN. This, in effect, up-regulates the
Figure 3. The fibrinolytic cascade. The cascade responds to FN
generated by the action of thrombin on fibrinogen. tPA is
released from the endothelium and, with FN as a cofactor,
catalyzes conversion of GPg to GPn, which begins to digest FN to
a modified form (FN⬘) with increased cofactor activity. FN⬘ also
serves as a cofactor for the plasmin-catalyzed conversion of GPg
to LPg, which is a better substrate than GPg for tPA. This
reaction and the conversion of FN to FN⬘ represent positive
feedback in the fibrinolytic cascade. When TAFIa is formed, it
further modifies FN⬘ to FN⬙ and thereby eliminates positive
feedback and attenuates fibrinolysis. The cascade is also down-
regulated by the serine protease inhibitors, PAI-1, and antiplas-
min (AP), which irreversibly inactivate tPA and plasmin, respec-
tively. TM ⫽ thrombomodulin.
Thrombin: Physiology and Pathophysiology
cofactor activity of FN⬘ to a level threefold higher than
that of intact FN. FN⬘ also serves as a cofactor for the
proteolytic conversions of GPg and GPn to their lysine
counterparts, lys-plasminogen (LPg) and lys-plasmin
(LPn). These reactions are catalyzed by GPn and LPn.
LPg is approximately 20-fold better than GPg as a sub-
strate for tPA-catalyzed formation of plasmin (LPn). Thus,
modification of FN to FN⬘ represents positive feedback in
the fibrinolytic cascade, somewhat akin to that repre-
sented by thrombin-catalyzed activation of factors V and
VIII in the coagulation cascade by thrombin. In response
to thrombin, and especially thrombin bound to thrombo-
modulin, TAFI is activated to TAFIa. TAFIa then cata-
lyzes removal of the carboxyl terminal lysine (and arginine)
residues present in FN⬘, thereby producing a form of
modified FN designated FN⬙.10 This down-regulates the
cofactor activity of FN with respect to both the activation
of GPg and the conversion of GPg to LPg. Thus, TAFIa,
while not eliminating plasminogen activation, eliminates
the positive feedback steps and thereby substantially
attenuates plasminogen activation and fibrinolysis. The
fibrinolytic cascade is also down-regulated by two fast-
acting serine protease inhibitors. One of these is plasmin-
ogen activator inhibitor type 1 (PAI-1), which targets tPA.
The other is antiplasmin, which targets GPn and LPn.
Interestingly, FN reduces the rate of inhibition of plasmin
by antiplasmin by a factor of approximately 100. The
fibrinolytic process terminates in the digestion of FN⬘ and
FN⬙ to a family of soluble FN degradation products
(FDPs), the smallest of which (DDE) is shown in Figure 3.
The coagulation and fibrinolytic cascades have similar
general features in common. They both respond to an
activator liberated from the tissues (tissue factor in coag-
ulation) or the endothelium (tPA in fibrinolysis), they both
proceed by zymogen to protease conversions, they both
have positive feedback steps and therefore generate ex-
plosive activity following an initiation phase, and they are
down-regulated by both protease inhibitors and reactions
mediated by the thrombin-thrombomodulin complex.
TAFI Protein and TAFI Gene
TAFI is a single-chain plasma protein of 401 amino
acids.12 Its peptide molecular mass is 45,999 d. In gel
electrophoresis in dodecyl sulfate, it migrates with an
apparent molecular weight of 60,000 d, a feature that can
be attributed to four glycosaminoglycan chains in the
amino-terminal region of the molecule.13 It circulates in
plasma at a concentration of approximately 100 nmol
(6 ␮g/mL), a value similar to that of protein C.14 The
plasma level of TAFI appears to be strongly determined by
polymorphisms in the promoter and 3⬘ untranslated region
of the gene.15 It is cleaved after arginine 92 by thrombin,
thrombin-thrombomodulin, and plasmin to yield a 92-
amino acid activation fragment from the amino terminus
and the carboxypeptidase B-like enzyme, TAFIa, from the
carboxyl terminus. The enzyme comprises 309 amino acids
and migrates in gel electrophoresis in dodecyl sulfate with
an apparent molecular weight of 35,000 d. The enzyme is
highly homologous to the pancreatic carboxypeptidase-
B.12 The enzyme, like that of the pancreas, is a Zn2⫹-
www.chestjournal.org
dependent carboxypeptidase and one molecule of zinc ion
has been found per molecule of the enzyme.16 The protein
was described by several groups at approximately the same
time, and therefore has acquired several monikers. It is
known as TAFI,14 plasma procarboxypeptidase B,17 pro-
carboxy-peptidase U (for unstable),18 and procarboxy-
peptidase R (for arginine).19 The chromosomal location for
TAFI is 13q14.11.20 The gene comprises 11 exons distrib-
uted on approximately 48 kilobase pairs of genomic
DNA.21 It is expressed in the liver. It also has been found
in platelets, and the messenger RNA has been found in the
megakaryocyte cell lines DAMI and CHRF.22
Activation of TAFI and the Inactivation
of TAFIa
TAFI is activated by thrombin alone at a rate and extent
sufficient to influence fibrinolysis, provided the thrombin
level is very high, such as that immediately following clot
formation in normal blood. In addition, it is very effec-
tively activated by thrombin at low levels in the presence
of thrombomodulin. Plasmin at high levels will also acti-
vate TAFI and the process is stimulated by heparin.23 The
physiologic activator, however, is presumed to be throm-
bin, and particularly the thrombin-thrombomodulin com-
plex.7 The kinetic mechanism for thrombin-thrombo-
modulin dependent activation is depicted in Figure 4. It
can be described as a parallel assembly, enzyme central
mechanism. In this mechanism, the central enzyme,
thrombin, can interact with either TAFI or thrombomodu-
lin to form the corresponding binary species. These then
further interact with the third component to form the
ternary thrombin-thrombomodulin-TAFI complex, from
which TAFIa is generated. TAFIa can also be generated
from the binary thrombin-TAFI complex, but the catalytic
efficiency of the ternary complex is 1,250-fold greater than
that of the binary complex. The reaction with thrombo-
modulin is Ca2⫹ dependent.14 This mechanism is formally
very similar to that described for protein C activation, but
numerous differences exist. More of the structure of
Figure 4. Mechanism of TAFI activation. TAFI can be gener-
ated from the binary thrombin-TAFI complex (T-TAFI) or the
ternary thrombin-thrombomodulin-TAFI complex (T-TM-
TAFI). Kd and Km values are the dissociation constants for the
corresponding interactions, and k1 and k2 are turnover numbers
for TAFIa generation. The catalytic efficiency for the thrombo-
modulin-dependent reaction is 1,250-fold greater than that for
the thrombomodulin-independent reaction.
CHEST / 124 / 3 / SEPTEMBER, 2003 SUPPLEMENT 35S
thrombomodulin is required for TAFI activation.24,25 In
addition, regions of thrombin that are sufficient for the
two reactions differ.26,27 Furthermore, the thrombomodu-
lin dependence of TAFI activation is not determined by
the amino acids surrounding the cleavage site.28 One
particularly notable difference involves methionine 388 of
thrombomodulin. This residue is highly susceptible to
oxidation, a reaction that can be accomplished by inflam-
matory cells.29 Oxidation of this residue reduces the
effectiveness of thrombomodulin in protein C activation
by approximately 90% but does not affect TAFI activa-
tion.24,29 This observation prompts the speculation that the
balance between FN deposition and FN removal might be
tipped strongly toward thrombosis in an inflammatory
milieu due to the loss of the anticoagulant but not
antifibrinolytic properties of thrombomodulin.
The enzyme, TAFIa, is intrinsically unstable.30,31 The
kinetics of decay are highly dependent on temperature. At
body temperature the enzyme has a half-life of approxi-
mately 8 min. The antifibrinolytic potential is directly
related to the half-life. Since no naturally occurring inhib-
itors of TAFIa have been found in the plasma, the intrinsic
instability appears to reflect the physiologic mechanism
for down-regulation. The region of the molecule associ-
ated with instability has been shown to encompass amino
acids 302–330.30,32 A common polymorphism in the hu-
man population has been found at position 325. One form
exhibits threonine at this position and the other isoleucine.
The latter form is twice as stable and 60% more potent as
an antifibrinolytic. Whether this fact has pathophysiologic
consequences, however, has yet to be determined, al-
though a recent report indicates that it is not associated
with a tendency for myocardial infarction.33
deficiencies in the initiation phase. Rather, they seem to
correlate more obviously with the failure to produce the
propagation phase. A logical conundrum still exists, how-
ever, because clotting is associated with the initiation
phase, which is nearly normal in hemophiliac blood.
Studies36 –38 have shown, however, that plasmas deficient
in components of the intrinsic pathway (factors VIII, IX,
and XI, but not XII), when clotted via the extrinsic
pathway, in the presence of tPA, exhibit subsequent
fibrinolysis three times earlier than normal plasmas. The
phenomenon has been designated “premature lysis.” The
mechanism behind this phenomenon has been worked
out. In normal plasma, the burst of thrombin following
clotting is sufficient, even in the absence of thrombo-
modulin, to activate sufficient TAFI to suppress subse-
quent plasminogen activation and fibrinolysis. This does
not happen in the hemophiliac plasma, however, because
of the lack of the thrombin burst in the propagation phase.
Consequently, the clots formed in the initiation phase
dissolve more quickly. The concepts are presented in
Figure 5. This premature lysis to date has only been
demonstrated in vitro. Its existence, however, has led to
the hypothesis that bleeding in patients with defects in the
intrinsic pathway may be as much due to the failure to
suppress fibrinolysis as it is to a failure to form a clot in the
first place.

Fibrinolysis and the Intrinsic Pathway of

Coagulation
A comprehensive analysis has been made of the tem-
poral course of events associated with the clotting of whole
blood when the process is initiated through the tissue
factor-dependent (extrinsic) pathway.34,35 The process be-
gins with an initiation phase, which lasts approximately
4.5 min. In this phase, thrombin at low levels catalyzes in
sequence platelet activation, factor V activation, factor
XIII activation, and FPA release. Over this interval,
the thrombin level reaches approximately 25 nmol
(2.5 U/mL). At the end of this phase, clotting occurs
(4.7 min). In blood from normal individuals, the initiation
phase is immediately followed by a propagation phase in
which the intrinsic pathway is triggered and a relatively
enormous burst of thrombin occurs within the clot, peak-
ing at a concentration of approximately 850 nmol. This is
then consumed fairly rapidly over the next few minutes by Figure 5. Suppression of fibrinolysis through the intrinsic
antithrombin and other protease inhibitors. In blood from pathway. Initiation of coagulation via the extrinsic pathway
generates small amounts of thrombin (IIa) from prothrombin (II)
people with deficiencies in the intrinsic pathway (eg, in the initiation phase. This is sufficient to form a clot by
hemophilia A or B), the initiation phase is similar to that of converting fibrinogen (FGN) to FN. It is also sufficient to initiate
blood from normal individuals, although clotting is de- the intrinsic pathway through the thrombin-catalyzed activation
layed modestly. Strikingly, however, the large burst of of factor XI. This leads to a large burst of thrombin formation
within the clot. The thrombin level is sufficiently high in this
thrombin following clotting is extremely diminished or phase to activate enough TAFI to suppress plasminogen (Pg)
absent. The dire bleeding tendencies in hemophilia A and activation to plasmin (Pn) and thereby attenuate digestion of the
B do not seem to be rationalized very well by modest clot to soluble FDPs.

36S Thrombin: Physiology and Pathophysiology

 Some Potential Clinical Implications of
the TAFI System
Insufficient time has passed since the discovery of the
TAFI system to have acquired enough information to
determine definitively whether it will be relevant with
respect to pathophysiology, diagnosis, or treatment. Epi-
demiologic studies and work in animals, however, suggest
that it might. For example, elevated levels of plasma TAFI
have been associated with a mildly elevated risk for venous
thrombosis,39 and it is significantly elevated (50%) in
patients with type 2 diabetes.40 In a large cohort studied in
France, an increase in the TAFI antigen level was found to
be a risk factor for angina pectoris; interestingly, the same
result was not seen in a study in Northern Ireland.41 TAFI
levels measured both by antigen detection and functional
assays were found to be significantly lower than normal in
disseminated intravascular coagulation.42 Similar measure-
ments in acute promyelocytic leukemia showed that anti-
gen levels were normal, but levels measured on the basis
of function were reduced approximately 60%.43 The sug-
gestion made in the report of this study was that the
acquired TAFI deficiency in acute promyelocytic leuke-
mia may contribute to the severity of the hemorrhagic
diathesis in this disease because of an impairment in the
capacity of the coagulation system to protect the clot from
the fibrinolytic system. Levels of TAFI were found to be
very low in patients with liver cirrhosis and even undetect-
able in patients with advanced hepatocellular disease.44,45
Recombinant factor VIIa, used for therapy of hemophili-
acs with inhibitors to factor VIII, was shown to prolong
clot lysis in vitro.46 The effect was shown to be TAFI
dependent. The authors concluded that recombinant fac-
tor VIIa both accelerates clot formation and inhibits
fibrinolysis in factor VIII-deficient plasma. They also
reported a large individual variability in antifibrinolytic
potential of recombinant factor VIIa in the population
studied. These studies and others like them all suggest that
TAFI may be linked to certain pathologic conditions, but
to date no pathologic condition has been identified that
can be linked directly to a defect in the TAFI system.
Studies in animal models show that TAFIa functions as
an antifibrinolytic agent in vivo. Early work by Redlitz et
al47 showed that a carboxypeptidase B-like activity is
induced when thrombolysis is initiated in a canine throm-
bolysis model. The activation of TAFI in another dog
model of coronary artery thrombosis was shown to be
inhibited by the thrombin inhibitor melagatran.48 Klement
et al49 investigated the impact of a TAFIa inhibitor on
tPA-induced clot lysis in a rabbit model of arterial throm-
bolysis. The inhibitor was a peptide isolated from the
potato called potato tuber carboxypeptidase inhibitor
(PTCI). They found that when the inhibitor was adminis-
tered along with tPA, numerous parameters associated
with thrombolysis were enhanced. The time to reperfusion
was reduced by an approximate factor of three. Vessel
patency was similarly improved. In their experiments, they
measured both clot accretion and lysis. The TAFIa inhib-
itor did not influence accretion but it substantially pro-
moted clot lysis. Similar studies were carried out by
Nagashima et al50 in a rabbit model of venous thrombol-
www.chestjournal.org
ysis, where the impact of PTCI on tPA-mediated clot lysis
was determined. Indwelling clots were removed and
weighted 90 min after initiation of the thrombolytic
treatment, which consisted of a bolus of tPA, with or
without PTCI, followed by continuous infusion of tPA,
with or without PTCI. They found that the weight of the
thrombus was not affected by PTCI when it was admin-
istered along with a saline solution control. When it was
administered along with tPA, however, the clot weight was
reduced to approximately one half of the control weight.
tPA alone at the same dose reduced clot weight only 26%.
This combined effect of tPA and PTCI could be obtained
with three times the dose of tPA in the absence of PTCI.
Thus, the TAFIa inhibitor effectively tripled the effect of
a given dose of tPA. In addition, when PTCI was admin-
istered along with the higher dose of tPA, the clots were
completely digested. A similar apparent tripling of the
effectiveness of tPA was reported by Refino et al51 in
another animal model of thrombolysis. These studies
suggest that TAFIa inhibitors could be fruitfully employed
as adjuvants in thrombolytic therapy.
Whether drugs will inhibit TAFIa or suppress TAFI
activation in order to promote endogenous fibrinolysis and
thereby offer long-term protection from undesirable
thrombotic events is not yet known. To date, no work in
animals along these lines has been reported. Most work in
vitro, however, indicates that TAFIa modulation affects
the balance between FN deposition and removal, and
thereby provides solid rationale for pursuing the appropri-
ate investigations. The TAFI knockout mouse has been
obtained.52 The mice are viable, and the absence of TAFI
in this species does not create a phenotype, which suggests
that TAFIa inhibition would not be unacceptably dan-
gerous.
In the case of hemophilia, a reasonable argument can
be made that an agent that would enhance TAFI activation
might help alleviate bleeding. An obvious choice for such
an agent would be a soluble recombinant thrombomodulin
engineered to promote TAFI activation but not protein C
activation. In vitro, soluble thrombomodulin has been
shown to correct the premature lysis of clots in hemophil-
iac plasma,36 because in the presence of thrombomodulin
the thrombin generated in the initiation phase is adequate
to activate TAFI, even in the absence of the thrombin
burst, which occurs in normal plasma during the propaga-
tion phase. Since the elements of thrombomodulin struc-
ture needed to activate TAFI and protein C differ some-
what, the acquisition of a form of thrombomodulin
selective for TAFI activation is feasible. In fact, two
potential variants are currently known. One with oxidized
methionine 388 is effective with TAFIa but not with
protein C.29 In addition, a construct with tyrosine 348
replaced with alanine functions normally in TAFI activa-
tion but not protein C activation.24
Whether measurements of TAFI, TAFIa, the inacti-
vated form of TAFIa, or other proteolytic fragments will
prove useful for diagnosis, prognosis, or risk assessment is
not yet known. This is also true for analyzing numerous
polymorphisms in both the translated and untranslated
portions of the TAFI gene. Several enzyme-linked immu-
nosorbent assays are available commercially, however. In
CHEST / 124 / 3 / SEPTEMBER, 2003 SUPPLEMENT 37S
addition, commercial assays exist that are based on the
measurement of carboxypeptidase B-like activity after
intentional activation of TAFI in the sample. A specific
assay for endogenous TAFIa, however, has proven difficult
to obtain. This is due, in large part, to the existence of a
constitutively active carboxypeptidase B-like enzyme
known as carboxypeptidase N.53 It is present at a relatively
high concentration in plasma and provides an unaccept-
ably high background activity for measuring TAFIa with
small and not sufficiently selective substrates. Our labora-
tory has devised an assay based on attenuation by TAFIa of
the plasminogen activator cofactor activity of high-molec-
ular-soluble FN degradation products. Preliminary work
has shown that endogenous TAFIa levels in normal plasma
are remarkably low, on the order of 5 to 10 pmol (the
TAFI zymogen level is approximately 100 nmol or 100,000
pmol). In collaboration with Dr. Fletcher Taylor, we have
found that continuous infusion of the baboon with throm-
bin at a low level over a period of 1 h leads to a progressive
increase in the plasma TAFIa concentration up to a level
of 2,000 pmol. On termination of the thrombin infusion,
the TAFIa level decreases approximately exponentially, with
a half-life of approximately 10 min. Thus, TAFI is activated in
vivo in response to thrombin, and the concentration of
TAFIa achieved is sufficient to seriously attenuate fibrinoly-
sis.14,31 Its relatively short half-life would suggest that ele-
vated TAFIa would reflect events that are currently happen-
ing, rather than those that happened in the past.
Summary
The coagulation and fibrinolytic cascades are generally
thought of as separate and independent entities. Recent
observations, however, suggest that they are directly
linked to one another. This occurs not only through the
TAFI pathway, but also through reactions whereby plas-
min, for example, can both activate and/or inactivate
coagulation factors V and IX.54,55 Consequently, methods
for more efficacious control of bleeding or thrombosis will
likely be obtained from insights gained through studies
aimed at acquiring a more global appreciation of the
mechanisms involved in regulating not just coagulation or
fibrinolysis, but rather in regulating the balance between
the deposition and removal of FN.
References
1 Davie EW, Fujikawa K, Kisiel W. The coagulation cascade:
initiation, maintenance, and regulation. Biochemistry 1991;
30:10363–10370
2 Mann KG, Bovill EG, Krishnaswamy S. Surface-dependent
reactions in the propagation phase of blood coagulation. Ann
N Y Acad Sci 1991; 614:63–75
3 Collen D, Lijnen HR. Basic and clinical aspects of fibrinolysis
and thrombolysis. Blood 1991; 78:3114 –3124
4 Irigoyen JP, Munoz-Canoves P, Montero L, et al. The
plasminogen activator system: biology and regulation. Cell
Mol Life Sci 1999; 56:104 –132
5 Giles AR, Nesheim ME, Herring SW, et al. The fibrinolytic
potential of the normal primate following the generation of
thrombin in vivo. Thromb Haemost 1990; 63:476 – 481
6 Esmon CT. Regulation of blood coagulation. Biochim Bio-
phys Acta 2000; 1477:349 –360
38S
7 Bajzar L, Morser J, Nesheim M. TAFI, or plasma procar-
boxypeptidase B, couples the coagulation and fibrinolytic
cascades through the thrombin-thrombomodulin complex.
J Biol Chem 1996; 271:16603–16608
8 Bajzar L. Thrombin activatable fibrinolysis inhibitor and an
antifibrinolytic pathway. Arterioscler Thromb Vasc Biol 2000;
20:2511–2518
9 Wang W, Boffa MB, Bajzar L, et al. A study of the mechanism
of inhibition of fibrinolysis by activated thrombin-activable
fibrinolysis inhibitor. J Biol Chem 1998; 273:27176 –27181
10 Nesheim M, Wang W, Boffa M, et al. Thrombin, thrombo-
modulin and TAFI in the molecular link between coagulation
and fibrinolysis. Thromb Haemost 1997; 78:386 –391
11 Collen D, Lijnen HR. Molecular and cellular basis of fibri-
nolysis. In: Hoffman R, Benz E, Shattil SJ, et al, eds.
Hematology: basic principles and practice. New York, NY:
Churchill Livingstone, 1991; 1232–1242
12 Eaton DL, Malloy BE, Tsai SP, et al. Isolation, molecular
cloning, and partial characterization of a novel carboxypeptidase
B from human plasma. J Biol Chem 1991; 266:21833–21838
13 Boffa MB, Wang W, Bajzar L, et al. Plasma and recombinant
thrombin-activable fibrinolysis inhibitor (TAFI) and activated
TAFI compared with respect to glycosylation, thrombin/
thrombomodulin-dependent activation, thermal stability, and
enzymatic properties. J Biol Chem 1998; 273:2127–2135
14 Bajzar L, Manuel R, Nesheim ME. Purification and charac-
terization of TAFI, a thrombin-activable fibrinolysis inhibitor.
J Biol Chem 1995; 270:14477–14484
15 Henry M, Aubert H, Morange PE, et al. Identification of
polymorphisms in the promoter and the 3⬘ region of the TAFI
gene: evidence that plasma TAFI antigen levels are strongly
genetically controlled. Blood 2001; 97:2053–2058
16 Marx PF, Bouma BN, Meijers JC. Role of zinc ions in
activation and inactivation of thrombin-activatable fibrinolysis
inhibitor. Biochemistry 2002; 41:1211–1216
17 Tan AK, Eaton DL. Activation and characterization of pro-
carboxypeptidase B from human plasma. Biochemistry 1995;
34:5811–5816
18 Wang W, Hendriks DF, Scharpe SS. Carboxypeptidase U, a
plasma carboxypeptidase with high affinity for plasminogen.
J Biol Chem 1994; 269:15937–15944
19 Campbell W, Okada H. An arginine specific carboxypeptidase
generated in blood during coagulation or inflammation which
is unrelated to carboxypeptidase N or its subunits. Biochem
Biophys Res Commun 1989; 162:933–939
20 Vanhoof G, Wauters J, Schatteman K, et al. The gene for
human carboxypeptidase U (CPU): a proposed novel regula-
tor of plasminogen activation; maps to 13q14.11. Genomics
1996; 38:454 – 455
21 Boffa MB, Reid TS, Joo E, et al. Characterization of the gene
encoding human TAFI (thrombin-activable fibrinolysis inhib-
itor; plasma procarboxypeptidase B). Biochemistry 1999;
38:6547– 6558
22 Mosnier LO, Buijtenhuijs P, Marx PF, et al. Identification of
thrombin activatable fibrinolysis inhibitor (TAFI) in human
platelets. Blood 2003; 101:4844 – 4846
23 Anderson GM, Shaw AR, Shafer JA. Functional characteriza-
tion of promoter elements involved in regulation of human B
beta-fibrinogen expression: evidence for binding of novel
activator and repressor proteins. J Biol Chem 1993; 268:
22650 –22655
24 Wang W, Nagashima M, Schneider M, et al. Elements of the
primary structure of thrombomodulin required for efficient
thrombin-activable fibrinolysis inhibitor activation. J Biol
Chem 2000; 275:22942–22947
25 Kokame K, Zheng X, Sadler JE. Activation of thrombin-activable
fibrinolysis inhibitor requires epidermal growth factor-like do-
Thrombin: Physiology and Pathophysiology
 main 3 of thrombomodulin and is inhibited competitively by
protein C. J Biol Chem 1998; 273:12135–12139
26 Hall SW, Nagashima M, Zhao L, et al. Thrombin interacts
with thrombomodulin, protein C, and thrombin-activatable
fibrinolysis inhibitor via specific and distinct domains. J Biol
Chem 1999; 274:25510 –25516
27 Bell R, Stevens WK, Jia Z, et al. Fluorescence properties and
functional roles of tryptophan residues 60d, 96, 148, 207, and
215 of thrombin. J Biol Chem 2000; 275:29513–29520
28 Schneider M, Nagashima M, Knappe S, et al. Amino acid
residues in the P6-P’3 region of thrombin-activable fibrino-
lysis inhibitor (TAFI) do not determine the thrombomodulin
dependence of TAFI activation. J Biol Chem 2002; 277:
9944 –9951
29 Glaser CB, Morser J, Clarke JH, et al. Oxidation of a specific
methionine in thrombomodulin by activated neutrophil prod-
ucts blocks cofactor activity: a potential rapid mechanism for
modulation of coagulation. J Clin Invest 1992; 90:2565–2573
30 Boffa MB, Bell R, Stevens WK, et al. Roles of thermal
instability and proteolytic cleavage in regulation of activated
thrombin-activable fibrinolysis inhibitor. J Biol Chem 2000;
275:12868 –12878
31 Schneider M, Boffa M, Stewart R, et al. Two naturally
occurring variants of TAFI (Thr-325 and Ile-325) differ
substantially with respect to thermal stability and antifibrino-
lytic activity of the enzyme. J Biol Chem 2002; 277:1021–1030
32 Marx PF, Hackeng TM, Dawson PE, et al. Inactivation of
active thrombin-activable fibrinolysis inhibitor takes place by
a process that involves conformational instability rather than
proteolytic cleavage. J Biol Chem 2000; 275:12410 –12415
33 Morange PE, Henry M, Frere C, et al. Thr325IIe polymor-
phism of the TAFI gene does not influence the risk of
myocardial infection. Blood 2002; 99:1878 –1879
34 Brummel KE, Paradis SG, Butenas S, et al. Thrombin
functions during tissue factor-induced blood coagulation.
Blood 2002; 100:148 –152
35 Butenas S, van ’t Veer C, Cawthern K, et al. Models of blood
coagulation. Blood Coagul Fibrinolysis 2000; 11(suppl 1):S9–
S13
36 Broze GJ Jr, Higuchi DA. Coagulation-dependent inhibition
of fibrinolysis: role of carboxypeptidase-U and the premature
lysis of clots from hemophilic plasma. Blood 1996; 88:3815–
3823
37 Mosnier LO, Lisman T, van den Berg HM, et al. The
defective down regulation of fibrinolysis in haemophilia A can
be restored by increasing the TAFI plasma concentration.
Thromb Haemost 2001; 86:1035–1039
38 Von dem Borne PA, Bajzar L, Meijers JC, et al. Thrombin-
mediated activation of factor XI results in a thrombin-
activatable fibrinolysis inhibitor-dependent inhibition of fibri-
nolysis. J Clin Invest 1997; 99:2323–2327
39 van Tilburg NH, Rosendaal FR, Bertina RM. Thrombin
activatable fibrinolysis inhibitor and the risk for deep vein
thrombosis. Blood 2000; 95:2855–2859
www.chestjournal.org
40 Yano Y, Kitagawa N, Gabazza EC, et al. Increased plasma
thrombin-activatable fibrinolysis inhibitor levels in normoten-
sive type 2 diabetic patients with microalbuminuria. J Clin
Endocrinol Metab 2003; 88:736 –741
41 Morange PE, Juhan-Vague I, Scarabin PY, et al. Association
between TAFI antigen and Ala147Thr polymorphism of the
TAFI gene and the angina pectoris incidence. Thromb
Haemost 2003; 89:554 –560
42 Watanabe R, Wada H, Watanabe Y, et al. Activity and antigen
levels of thrombin-activatable fibrinolysis inhibitor in plasma
of patients with disseminated intravascular coagulation.
Thromb Res 2001; 104:1– 6
43 Meijers JC, Oudijk EJ, Mosnier LO, et al. Reduced activity of
TAFI (thrombin-activatable fibrinolysis inhibitor) in acute
promyelocytic leukaemia. Br J Haematol 2000; 108:518 –523
44 Van Thiel DH, George M, Fareed J. Low levels of thrombin
activatable fibrinolysis inhibitor (TAFI) in patients with
chronic liver disease. Thromb Haemost 2001; 85:667– 670
45 Lisman T, Leebeek FW, Mosnier LO, et al. Thrombin-
activatable fibrinolysis inhibitor deficiency in cirrhosis is not
associated with increased plasma fibrinolysis. Gastroenterol-
ogy 2001; 121:131–139
46 Lisman T, Mosnier LO, Lambert T, et al. Inhibition of
fibrinolysis by recombinant factor VIIa in plasma from pa-
tients with severe hemophilia A. Blood 2002; 99:175–179
47 Redlitz A, Nicolini FA, Malycky JL, et al. Inducible car-
boxypeptidase activity: a role in clot lysis in vivo. Circulation
1996; 93:1328 –1330
48 Mattsson C, Bjorkman JA, Abrahamsson T, et al. Local
proCPU (TAFI) activation during thrombolytic treatment in a
dog model of coronary artery thrombosis can be inhibited
with a direct, small molecule thrombin inhibitor (melagatran).
Thromb Haemost 2002; 87:557–562
49 Klement P, Liao P, Bajzar L. An inhibitor of activated TAFI
enhances thrombolysis. Blood 1998; 92:1967–1975
50 Nagashima M, Werner M, Wang M, et al. An inhibitor of
activated thrombin-activatable fibrinolysis inhibitor potenti-
ates tissue-type plasminogen activator-induced thrombolysis
in a rabbit jugular vein thrombolysis model. Thromb Res
2000; 98:333–342
51 Refino CJ, Schmitt D, Pater C, et al. Carboxypeptidase
inhibitor markedly improves the potency of tPA in vivo.
Fibrinolysis Proteolysis 1998; 12:29
52 Nagashima M, Yin ZF, Zhao L, et al. Thrombin-activatable
fibrinolysis inhibitor (TAFI) deficiency is compatible with
murine life. J Clin Invest 2002; 109:101–110
53 Skidgel RA. Human carboxypeptidase N: lysine carboxypep-
tidase. Methods Enzymol 1995; 248:653– 663
54 Lee CD, Mann KG. Activation/inactivation of human factor V
by plasmin. Blood 1989; 73:185–190
55 Samis JA, Ramsey GD, Walker JB, et al. Proteolytic process-
ing of human coagulation factor IX by plasmin. Blood 2000;
95:943–951
CHEST / 124 / 3 / SEPTEMBER, 2003 SUPPLEMENT 39S